ECW12
Office Futures
ECW12
Issue 12, June 14, 2000
+++ Ecommerce standards: why they are needed and what some of them do
+++ Rant: What have we learned?
+++ We get mail: Frank Glyn-Jones
+++ Ecommerce standards: why they are needed and what some of them do
++ What makes a standard succeed?
There is nothing new about electronic commerce. Companies have been exchanging trade data over telecommunications networks almost since the beginnings of electrical telegraphy, nearly 150 years ago. Since then, numerous other methods have arisen for 'telegraphy' (Greek for 'writing at a distance').
Almost all have been enthusiastically adopted by business. Technologies such as Telex, videotex, facsimile transmission, electronic mail and, most recently, the World Wide Web have each, in turn, created a widely-used business communication channel. In doing so, they have also created a mass market for system and service suppliers.
Which is chicken and which is egg in each case is hard to determine. Clear in every one, though, is that establishing the necessary critical mass of users has involved a third force -- external, often governmental, regulation. Supply push and demand pull cannot do it alone. It has needed all three forces acting in combination.
Where any one of these ingredients has been lacking, worldwide adoption has proved elusive. AT&T's TWX (teletypewriter exchange) system, for example, had both supply push and demand pull within the USA. It was, though, incompatible with the Telex standards defined by the CCITT, the international telecommunications body, and was therefore little used outside the USA and its dependencies. (In return, the USA largely shunned Telex. When Western Union bought TWX in the early 1970s, it offered its American customers a merged system but there was little interest.)
An example of a different lack was provided by Teletex, a kind of super-Telex. In the early 1980s, the European PTTs (telecommunications authorities) were trumpeting this as the next big thing, likely to oust both Telex and videotex. It got nowhere. The high cost of user equipment, the imminence of the X.400 electronic mail standard and some stunningly incompetent marketing and tariff-setting made it an orphan. Standards compatibility was there but not the other two ingredients. By the end of the decade, Teletex was effectively dead.
These three factors -- supply push, demand pull and external regulation -- are applicable also to Internet electronic commerce. If it is to become an open, worldwide trading channel, it must also have this seemingly inescapable trio of elements in place.
On the face of it, there seems no reason why a proprietary method or a de facto standard should not provide the necessary control but there is no historical precedent for this. Not even in the days of AT&T's complete dominance of the American telephony market were its technical methods adopted worldwide. Microsoft is probably hoping to be the exception that proves this rule (see Biztalk, below), but is finding that harder than in the past.
++ What standards make possible
To illustrate why standards are needed, one need look no further than some of the current leaders in Web-based trading. Amazon.com, Intel, Cisco and Dell, for instance, are all presently enjoying high-profile success with their commercial Web sites. These companies' use of Internet technologies is, quite reasonably, being held up as illustrations of what is possible.
Admirable as the achievements of these companies may be, they need to be seen for what they are. Each operation is a 'one-off', employing protocols and methods that are not in widespread use and that are often of the company's own devising. Their electronic commerce operations are, in a sense, freaks. (Freaks born of inspiration, hard work and doggedness, without a doubt, but abberations all the same.) None has used standard, or even especially similar, Web-trading methods.
This is not to criticize Amazon.com, Intel and the others. There were no standard methods available to them when they began Web trading. There are virtually none even now. It is merely to point out that, while their achievements are worth trying to emulate, their technical methods are not necessarily suitable for or even usable by other organizations. Any firm wanting to follow Amazon's example, for instance, therefore needs to think carefully about the implications for itself and its customer and trading partners before embarking on electronic trading.
Trading implies a relationship. At its most basic, this consists of a single buyer dealing with a single seller. If matters were always that simple, there would be little need for external standards, especially if one or other were dominant. The weaker partner would simply have to do things the other company's way. The same is largely true if a dominant buyer deals with a number of sellers. Anyone supplying organizations like car makers, supermarkets or national governments will know the truth of that. (Even then, the coverage is partial. For instance, only about 60 per cent of Tesco's suppliers trade with it electronically.)
If, however, each of those buyers and sellers deals with other buyers and sellers, the problem becomes anything but simple. Unless the whole network of relationships is controlled by the one organization, which is very rare, there are two main outcomes likely.
The first, and better, outcome is that common standards are created or adopted. This helps these various trading entities achieve some measure of control and efficiency. It may also lead to qualitative improvements in the nature of their relationship.These standards could be de jure, such as those widely used in Electronic Data Interchange (EDI). Alternatively, they could be de facto, such as might be made available by a third-party services supplier. General Electric's GE Information Services (GEIS) does both. It is the dominant EDI services supplier in Britain, numbering the top ten supermarkets among its 7,000 UK customers. It has its own methods, which it offers traders, as well as those based on defined standards or those of the trader's customers.
The other possibility in a complex trading network is that each buyer-seller pairing makes its own rules and formats. This way lies delay, error and huge inefficiencies. Ask almost any book publisher or, until recently, anyone who supplies parts or materials to more than one car maker.
++ A sample value chain
To illustrate some of the issues involved, here is a simple trading network. Imagine that a trading relationship exists among the four organizations listed earlier. In this, Intel, as an electronics components manufacturer, regularly does large volumes of business with Cisco, the maker of computer networking equipment. Further, Cisco regularly trades in quantity with Dell, maker of small to medium-sized computers.
Finally, assume that Amazon.com, the online bookshop, has an arrangement with Dell to sell manuals and text-books associated with the latter's products. (In fact, Dell's Web site already has a 'passthrough' feature on it, leading straight to Amazon's site..)
Thus, there is a simple value chain running from Intel to Dell, via Cisco. Amazon.com is off to one side but is linked to Dell, as shown below.
Intel . . . . . . .Cisco. . . . . . . . Dell
|
|
Amazon.com
In this fictitious scenario, Intel does not deal direct with Dell. The reality, of course, is that each of the three manufacturing companies buys direct from a range of component suppliers. Margins in the electronics sector seldom allow for one-to-one supply chains like this, but they can be found in other sectors.
++ Problems arising along the chain
What if these four companies now wanted to consummate this trading relationship electronically? Which methods and formats would and could they use?
In answering those questions, their respective systems teams would need to deal with the following ten issues, among others.
1. What ordering and shipping system must each company use in dealing with its trading partner or partners? How will orders, receipts, part-shipments, variations, returns and so on be dealt with? If some or all of them are using the same logistics company, can their shipping systems be unified or at least made similar? Will Cisco, in the middle, have to use two alien systems, as well as its own?
2. What payment system must each company use in dealing with its trading partner or partners? How will invoices, statements, remittances and so on be dealt with? (Cisco has to look two ways here, too.)
3. What payment systems must a trade customer use if buying, at various times, from all three manufacturers -- Intel, Cisco and Dell? Must it fall into line with each of three different requirements?
4. How will an occasional purchaser buy from Dell if he also wants some computer manuals from Amazon.com? Which company's payment method must he use for the books? If it is Dell's, then how will the payment make its electronic way to Amazon.com? If it is Amazon.com's, will the buyer be irritated by having to input two lots of credit card and personal details for what he sees as one transaction?
5. With any of the four companies, how does the occasional purchaser know he is, in fact, dealing with that organization and not someone masquerading as it? Similarly, how will any of the four companies know that the purchaser is who says he is? Will the same authentication method suffice in each case and in each direction?
6. As important, if not more so, how can Intel and Cisco be certain that the person ordering the next 50,000 components or 5,000 sub-assemblies does indeed work for the next company up the line? How will they know whether he is authorized to place that amount of business? And will the methods used for this be the same as for each company's other trading partners (who may number in their hundreds)?
7. Even if the end-points have been verified, how can these three trading partners keep the details of their orders secure from unauthorised view? Do they encrypt the messages passing between them and, if so, can they all use the same method? Will this be allowed, to the same level, by the regulatory authorities in all the countries they operate in?
8. What do the companies do with the marketing information they have gleaned from their commercial activities? Will they be swapping customer details, such as names, addresses, shipping information and credit experience, either locally and across national borders? Will they be allowed to do so by the relevant regulatory authorities?
9. What taxes and duties need to be paid on an order for a Cisco router, say, being bought from Dell by a French customer? In our imaginary supply chain, the components will have been shipped from one of Intel's factories in the USA to Cisco's plant in California. Will there be any inter-state taxes to pay?
The resulting device needs further work before sale, say, so is then forwarded to Dell's European manufacturing plant in Ireland. From there, it is shipped to France. Do these three countries -- the USA, Ireland and France -- have a consistent approach to tariffs on electronic trade? Would it be cheaper for our French buyer to import direct from Cisco, for instance (thus paying duty only once), and doing the final assembly work himself?
10. Finally, if the buyer in France were so unhappy with the performance of the router bought from Dell that he sought redress in court, where would the case be heard? Was the contract made in France, where he is, or in England, where Dell's European headquarters are?
The first six of these issues clearly relate to technical or computing considerations, such as defining data sets and document formats and aligning processes. The remainder are mainly or completely to do with marketing, regulatory and legal matters. Although these are outside the control of the individual organizations, their limiting or liberating effects should not be underestimated. At the moment, the Internet is only lightly policed. The world's governments and tax-gatherers will not be letting that situation continue much longer. Nor will lawyers.
It is at the system level where these four companies have greatest control over their actions. If there were standards in place for some of these aspects, life would be that much easier for our four companies. Progress would be achieved by adoption and adaptation, rather than by invention. Since the three manufacturing companies are of broadly similar size, there is no question here of one partner dominating. True collaboration is needed.
++ Electronic Data Interchange (EDI)
The few standards that are available at the moment are mostly to do with Electronic Data Interchange, either X.12 in the USA or EDIFACT in Europe. As EDI's name suggests, it is concerned mainly with data movement, not with process management. Thus, in terms of our ten sets of issues, it can deal with number 1 (ordering and shipping systems within the chain) but only in part. EDI schemes typically also include provision for items 6 and 7 (identification and authorization within the chain and the security of trading data).
Because EDI is built on a messaging model, usually running over private networks, it needs to be bridged to the worlds of real-time or Web-based working. Doing this requires either special software, which is only now becoming available, or a suitable service. GEIS's TradeWeb service, for instance, gives subscribers browser-based input, over the Internet, to their trading partners' EDI systems. These companies need also to be using GEIS's services.
There is much activity on the technical standards front. Various groups, having recognised the need for solutions to the problems like those described here, are now starting to make public their proposals. These are in varying stages of development. Here are some of the more significant ones.
++ Collaborative Planning Forecasting and Replenishment (CPFR)
This deals with some of questions 1 (ordering and shipping systems within the chain) and 8 (exchanging customer information). It is being put forward by the Voluntary Interindustry Commerce Standards Council (http://www.vics.org), an American group whose members include Nabisco Food, Wal-Mart, Procter & Gamble and Kmart.
VICS' intent for the standard is for it to help 'create unified business rules among multiple trading partners' and to manage sales data. It therefore deals with the issues of ordering and shipping, and with exchanging marketing information.
CPFR is designed to work into existing EDI protocols and will also be compliant with object-oriented standards, such as Microsoft's COM (Component Object Model) and the Object Management Group's CORBA (Common Object Request Broker Architecture), and also with XML (Extensible Markup Language). Several software suppliers are incorporating CPFR methods in their products, including Syncra Systems, Logility, Manugistics and i2 Technologies.
Although it is, at the moment, optimized for retailers, CPFR is intended to work across industry sectors. It looks likely, in due course, to be exactly what our three manufacturers are looking for on those issues.
But is 'in due course' soon enough? This is the problem with complex standards like these. Work on CPFR began in late 1996, with a pilot study being conducted the following year. It was not until June 1998 that VICS approved and published the CPFR guidelines (see http://www.cpfr.org). Only now, n
early two years later, are suppliers beginning to offer compliant products and traders to adopt them in their production systems.
In the meantime, other traders have been devising their own solutions. In the USA, Heineken had already created its own, in 1995, called the Heineken Operational Planning System (HOPS). In Britain, Safeway, followed by Tesco, introduced bespoke collaborative systems in 1998.
These companies have forsaken potential interchangeability for a head start of at two years or more (five, in Heineken's case). It is the sort of calculation that all Web traders will have to make, whether on this or other interchange standards.
++ Internet Open Trading Protocol (IOTP)
This is intended to unify the various electronic payment schemes presently on offer. IOTP was originally put forward, as OTP, by the Open Trading Protocol Consortium, which included IBM, BT, Netscape, Sun Microsystems, Mastercard, Mondex and Wells Fargo Bank among its members (but not Microsoft or Visa).
Work on it was taken over by the Internet Engineering task Force (IETF), thus removing it from proprietary hands. Almost inevitably, this has slowed progress. The OTP Consortium's first drafts date from 1998; now as IOTP it was about to move from an Internet Draft to an RFC (Request for Comment) in April 2000. Commercial adoption is therefore still some way off.
IOTP works with a range of payment systems, including DigiCash's ecash, Mondex, SET (Secure Electronic Transactions), e-checks and debit cards. It will also work with micro-payment systems. These allow electronic 'purses' (virtual petty-cash boxes) to be filled at intervals, usually from a credit card, and drawn from to make small payments.
All parties concerned in a transaction will be able to use IOTP-based systems, whether customers, merchants, credit checkers and certifiers, banks or 'delivery handlers'. (In its original form, OTP, it excluded consumer aspects.) The standard uses Extensible Markup Language (XML) to define data, which it transfers in a message format. Also, it can incorporate digital signatures.
The objective for IOTP is to make electronic purchase transactions consistent for all involved, regardless of payment system. It thus deals with questions 2 and 3 (payment systems within and outside the chain) and some of questions 4 and 5 (payment systems for and identification of occasional purchasers). It also meets some of the concerns in question 6 (identification and authorization within the chain).
++ Open Buying on the Internet (OBI)
This is the work of a different consortium from that behind OTP (as was). Its supporters include Microsoft, Visa, Oracle, Barnes & Noble, Lockheed Martin, SAP America, Texas Instruments, IBM (again), as well as several financial organizations. Chief among these is American Express, who helped underwrite the work on the standard, which is now released.
OBI is intended for 'high-volume, low-dollar transactions' between businesses (see http://www.openbuy.org for more details). It seeks to standardize:
- the process by which a requisitioner accesses an online catalogue at a selling organization
- the data format for order-related information exchanged between trading partners
- the method(s) for transmitting order-related data between organizations, and
- the security mechanisms involved in all these.
Thus, OBI promises to meet the needs expressed in items 2 and 3 (trade payments within and from outside the chain) and item 6 (trade buyer authentication) of the issues list. When it will do this is unclear; at the time of writing, it was in draft version 2.1. In July 1999, Ford Motor Company purchased some medical supplies for its company infirmary using OBI-compliant software from Intelisys.
Like IOTP, OBI incorporates other standards, among them the US X.12 EDI standards, Secure Sockets Layer, and digital certificates based on X.509. It does not, though, use XML, being based on HTML for data definitions. This will lead to obsolescence and incompatibilities unless it is reworked to do so. It does not work with IOTP.
++ BizTalk
This is even more proprietary than OBI, being wholly of Microsoft's invention, but is narrower in focus. It also is slow in appearing, having been announced in March 1999.
BizTalk consists of a product and a technology. The product is BizTalk Server, a program for linking organizations' e-commerce systems. Microsoft announced the 'preview release' of this in April 2000, the real thing being expected "later this year". Part of the delay has been because the software runs within Windows 2000, using its Active Directory, and this was itself late in shipping.
BizTalk Framework is the technology. It is a set of guidelines for companies to publish definitions ('schemas') of electronic commerce documents in XML. It also includes methods for using XML messages to integrate software programs. Compliant documents will be included in a publicly accessible repository (see http://www.microsoft.com/biztalk ). At this stage, it looks likely that the BizTalk Framework will deal with some of question 1 (ordering and shipping systems within the chain) and question 8 (exchanging customer information).
The framework already has support from SAP, J.D. Edwards and Peoplesoft; a 'starter kit' CD for it became available in December 1999. Because it is a Microsoft standard, some people are looking at it askance. Others, usually those already committed to using Microsoft methods, welcome it. Dell, for example, is using an early version of it already.
There is competition to Biztalk, not only from de jure standards like IOTP but also from other suppliers, such as IBM (with tpaML, or Trading Partner Agreement Markup Language) and Hewlett Packard (with 'e-speak', aimed mainly at e-commerce services).
++ Electronic Business XML Initiative (ebXML)
To complicate matters further, there is the Electronic Business XML Initiative (ebXML). This is "an open, vendor-neutral initiative to establish a global technical and semantic framework that will enable XML to be used in a consistent manner for the exchange of electronic business data".
ebXML is thus yet another de jure standard in the making. It has the support of the United Nations and of OASIS (Organization for the Advancement of Structured Information Standards. See http://www.oasis-open.org). All the major computer and software suppliers belong to OASIS, including Microsoft, IBM and Hewlett Packard (who are all doing their usual stunt of riding at least two horses at once). Another OASIS member is Visa, which has announced an XML-based set of invoicing protocols that will comply with ebXML.
The ebXML standard is not yet finalised but it looks likely to cover at least the same ground as IOTP and also to deal with question 1 (ordering and shipping systems within the chain). The bodies behind the two will therefore need to announce a convergence strategy pretty soon if they are not to leave Microsoft to benefit from the uncertainty.
++ Platform for Privacy Preference (P3P)
Current activities on standards for privacy and data protection are likely to bear on the technical aspects, at least, of issue 8 (exchanging customer information). The major programme in progress is the Platform for Privacy Preference Project (P3P), an initiative of the World Wide Web Consortium (see http://www.w3c.org/P3P). It is intended to 'enable computer users to be informed, and to make choices, about the collection, use and disclosure of their personal information on the Web'.
The intention is to establish a set of privacy-related terms and a protocol for using them, allowing Web browsers fitted with P3P software to interact with P3P-enabled Web sites. It would work much like the cookie alert feature in Microsoft's Internet Explorer, with user-definable settings. The software will put up a warning window if someone goes into a site whose data-harvesting policies are at variance with the browser settings.
P3P has been no faster in arriving than the other vendor-neutral standards discussed here, having taken about three years to reach final draft status. At this stage, there is no certainty about which suppliers and traders will implement it, nor which laws it will have embodied in it.
++ Summary
The application of the various standards and methods described above is summarized here.
1. Ordering/shipping within chain: EDI (in part), CPFR, BizTalk (in part) and ebXML
2. Payment within chain: IOTP, OBI and ebXML
3. Payment -- regular purchasers from outside: IOTP, OBI and ebXML
4. Payment -- occasional purchasers from multiple vendors: IOTP (in part) and ebXML
5. Identification of and for occasional purchasers: IOTP (in part) and ebXML
6. Identification/authorization within chain: EDI (in part) and OBI
7. Security of trading data: EDI (in part), IOTP (in part) and OBI (in part)
8. Exchanging customer information: CPFR, BizTalk, ebXML and P3P
9. Taxes and duties: none
10. Legal redress: none
The entries for item 7 (security and identification) reflect the fact that these particular schemes sensibly make use of existing security standards and systems, such as X.509 and SSL.
From this brief analysis, one can see that:
- there is plenty of overlap, there being at least two contenders for each of the first eight issues (the last two, as we have discussed, are not amenable to technical solution)
- two of those contenders are open, vendor-neutral standards -- IOTP and eXML. The sooner they are merged the better
- Microsoft has a direct or controlling interest in two schemes -- OBI and BizTalk -- that cover the whole eight issues
- finalization of most of these standards is months away, their implementation even further.
The output of the various committees and consortia working on these schemes is vital to the widespread use of e-commerce. While haste, possibly resulting in error, is not wanted by anyone, a greater enemy would be undue delay in producing acceptable and workable standards. The next few months are not only crucial, they should see a greater clarity in what is available and their exact areas of coverage.
-oOo-
+++ Rant: What have we learned?
Roger and I spent most of last year playing Cassandra, saying that there would be tears before bedtime. As the Dot.Coms crash around the ears of the Gaderene swine, the pearls are being thrown out with the bathwater.
Or something like that. Mix your metaphors as you will [That's mixing? More like giving them ten minutes in the kitchen blender! -- Roger], who has really got hurt and what has really changed? Sadly, the answer is "Almost none of the guilty" and "Almost nothing".
A few instant millionaires became somewhat less affluent, quite a few very foolish day traders got blown out of the water and a lot of employees now sitting on a pile of worthless stock options are keenly seeking new avenues to instant wealth. But the investment bankers who took huge fees for placing the issues are still charging huge fees for the next generation of hopefuls; the big accountants who did the due diligence and business plans are getting fat fees for bankruptcies and restructurings. The vulture capitalists are still out with their road shows, now pooh-poohing e-commerce, but snapping up wireless, optical and all the toys which may or may not power tomorrow's go-faster Internet.
Cybermalls are out. Portals and vortals are out. Pet shops are out. But already received wisdom is coalescing around tomorrow's winners. In Economics 101, they teach you that markets are cyclical. In the new Internet Age, the cycles are just faster. Stupidity, however, seems built into the system. This must be some sort of golden age for contrarians.
If we pretend we are men from Mars landing on earth for the first time, what do we see when we look at the Internet?
While we are impressed with the speed with which the Web has spread (although just over half the speed of diffusion of radio in the US), we are puzzled that what looks like a cheap communications system is most highly concentrated where affluent people already enjoy excellent and relatively cheap communications. We are surprised that some marketing geniuses have convinced most users that they have to spend a multiple of what they really need to get a machine capable of putting them on line. We note with surprise that while hundreds of millions are being spent to direct users to high cost commercial sites, the most intensive use of the Web seems directed towards free mail, free music, free software, free chat and free porn.
And our Intelligence Officer asks just why it is that the people who seem to be guiding all the big money deals are the same people who have been wrong on pretty much every IT issue over the last 20 years. It is as if the monks in the scriptorium had been given responsibility for diffusing Gutenberg's precocious child.
The interesting outcome is that, effectively, we have two Internets. There is the big money big engine internet of the telcos and traditional content providers, the investment bankers and big IT consultancies. On that internet, there is the promise that a FTSE star of 1999 can be assured of being a FTSE star in 2009 -- just listen to our received wisdom and spend lots of money.
Then there is the Wild West internet of the geeks and freaks, the playground of the mad, the bad and the sad, the Bizarre Bazaar where $1000 software is free if you can run ARJ and open an .nfo file, where you can take your pick of hundreds of thousands of pieces of CD-quality music, where every perversion imaginable (and some still unimaginable even when you stumble across them) is catered for.
The former is a stately structured place, where your guides wear thousand quid suits and assure you of their expertise, surrounded by mahogany and chrome and insulated with seven and eight figure fee notes. The latter is a place of ramshackle excellence where hackers and crackers plot mad schemes to deface the CIA site, break open DVD security systems and generally give anarchic effect to their proposition that "information wants to be free".
---------------snip----------------------------
(There's nothing particularly new about such thoughts.
Here's an extract from Roger's 1995 opus on workgroup computing)
The Internet itself, with its 'free' services, is representative of
this [sharing approach]. This will not last. The increased speed,
reliability and security that commercial users will demand will force
either the transformation of the Internet into a pay-as-you-go service
or else the creation of a second Internet. This would most likely be
run by the existing capacity providers - the telephone companies,
broadcasters, cable television companies and the on-line service
suppliers - and would have gateways into the 'old' Internet. What is
today called the Internet may well remain as the electronic equivalent
to the public transport system - cheap and companionable but erratic.
------------------------/snip-----------------------
There is surprisingly free movement beyond the two worlds. The odd-looking young men that the gentlemen in expensive suits keep decently hidden in back rooms often meet with co-conspirators in plotting midnight attacks on the establishment. The wild-eyed lunatic running an ISP in the backwoods with a Dodge half-ton and a quart of shine is a reserve member of the National Security Agency, available for call-up at any time.
Probably not since the time of the Elizabethans -- when a respectable merchant could turn buccaneer then pirate then merchant again a dozen times on a single voyage -- has the dividing line been so ill-defined.
Learning the right lessons is crucial. Most IT companies were desperately wrong about personal computers twenty years ago. They were not stupid people. They were certainly no more stupid than the people running passenger steamships in the 1950s, railroad systems at the dawn of the internal combustion engine, or the telegraph moguls when Mr Bell's quaint system for sound over wire came along.
It is human nature to try to cast the new into the shape of the old. Roger and I had fun at the expense of beancounters and that noble institution, Barclays Bank, a few years ago when they thought that e-commerce would take the shape of an electronic mall.. Of course, telcos want to make the world of packet switching operate on their expensive switched circuit. Copper owners will try to persuade you that asymmetric is logical, because that's how their systems operate. The banks would rather like to charge you a premium for the privilege of saving them money by banking on the Internet. And it's entirely logical that the credit card companies would think it's pretty neat that they charge more for electronic commerce. Most of all, it is human nature to want to own a nice chunk of tomorrow's big winner today.
The odd thing is that we are prepared to bet on the advice of people whose primary thought process is that tomorrow's outcome can most accurately be assessed on the basis of hundreds, thousands or even millions of outcomes yesterday.
Looking backwards is sensible. We do it all the time. But you have to look to what happened in other periods of disruptive change. In other periods of disruptive change, the established order stampeded like a herd of cattle into Tulip Mania and the South Sea Bubble. You can take it as gospel that when that unsung genius first started a fire by striking a flint, a thousand village elders bet the whole tribal stock of cowrie shells on schemes to achieve the same with granite, chert, basalt . . .
The crucial thing to remember is that when the herd moves, it is the elders -- the custodians of a thousand yesterdays -- who are doing the counselling.
What have we learned?
When it comes to disruptive change, the old rule of the social sciences is doubly valuable: don't listen to what people say, watch what they do.
Company A brags that all its employees are on an intranet. Smile politely and ask for the log files. Ask who may post content. Find out whether all the action instead really happens on informal group intranets of which the management knows nothing.
Company B tells you they have an industry portal which "positions them for international leadership". They tell you they spent $10 million with a top international consultancy. Ask what user research they have done, what are their repeat visit rates and average length of "visit". Check for real time interactivity and connectivity. Be gentle.
ING tells you they have introduced humanity and efficiency into internet banking. Believe them. It can be done.
Yes, there are winners out there. It just isn't necessarily the folks you'd expect or the ones spending the big bucks to say it's them.
Mostly, it's a world of Bourbons: people who have learned nothing and forgotten nothing. And, of course, there are the communards storming the barricades.
If you are going to rely on received wisdom -- and most of us have to most of the time -- just make sure you're listening to someone who doesn't believe that the history of technology and economics began in 1945 [or, if they're under thirty, 1985].
-oOo-
+++We get mail: Frank Glyn-Jones
I like your vision of what Internet will achieve. You talk about it levelling out prices, cutting out the middle men, level playing fields for rich and poor, and affecting Government /people relationships. I presume that you believe, as I do, that the Net is in its infancy and that we will in due course see broadband with huge increases in Net efficiency. Costs will come down; Net access will spread like the phone with over 90% of households in the West having access. We will see videoconferencing becoming commonplace and also the electronic book and Internet telephony.
All these things are likely to have a major impact. For example, Internet telephony could have a terminal impact [no pun intended] on BT and other European telecomms. BT and the Government might try to block it by taxation or licensing but I doubt if they will succeed.
I'm not up to date on machine translation. Is that possible? My own experience with translation over the Net indicates that it is hopeless. Anyway, the fact is that the Net is the focus of much of the world's R&D investment effort, which is accelerating.
I believe that cyber-learning will become one of the really big applications and will have an impact on skills. Many believe that the next revolution in human evolution will be the skills revolution and I believe the Net will be at the heart of that. It has started in the US and we are badly behind in this.
I do work on the future of Eastbourne, which has one of the lowest average incomes per head in the south east [of England] because the road and rail systems are so bad that business will not come down here. But if we had video-conferencing and broadband communications, that might change radically
.
My overall view of the future is as follows and I would greatly appreciate your [and other readers'] comments. I think the key factor is the availability of risk finance. It is new start-ups that are shaping the future. One reason why the US is so good is that they have evolved a really efficient venture capital system. Have a good VC system and that in itself encourages innovation. Of course, there will be failures but overall the system works.
We in the UK are much more risk averse and the system is even worse in Europe. This is connected with the high tax and regulatory regimes. The Far East is much better placed and is coming to life. In both the US and Asia government spending is much lower and hence more money is available for risk finance.
Of course, start-ups have to get big in order to affect the overall economic scene but they are doing just that in the US. If you look at a list of the top 100 IT companies in the world most are in the US and most are fairly new, e.g. Amazon, eBay, Netscape, Cisco, AOL and so on.
I see Net evolution speeding up and, if we want to play a significant role in it, we [the UK] need to have much more venture capital and fewer regulations and that means a culture revolution. We need to get public spending down to 35% or less, releasing about GBP40 billion for venture capital. We have to abandon the welfare mentality and go for growth. We have a few growth firms, i.e. Vodaphone and Orange, but not too many. We have a lot of start-ups, 700,000 last year, but not many get to the growth phase.
To grow they have to become global and these days this does not mean exports, it means setting up overseas affiliates. IBM is not a big exporting company, it has a world network of affiliates, as do ICI and the other multinationals. Multi-nationals account for an ever increasing part of world production. When Japan was in the growth phase during 1960-1990, it was spawning a large number of huge firms, Sony, Hitachi, Honda, etc. and about 12 giant companies accounted for most of Japan's growth. Same with US today. So we have to generate growth companies on the global scale, hence the need for a cultural revolution. Ministers' response is to say, "Yes, we agree that public spending has got to come down, but tell us where the cuts are to be made, health, education, pensions, etc". And, of course, as a society we are demanding more public spending on these every year.
I'm saying let's look at it differently. If we don't go for growth, market forces will cut public spending for us.
Anyway, let's stop there. Please let me know what you think.
frank.glyn-jones@virgin.net
(Frank is an independent analyst and journalist. At one time he was a member of IBM's Office Products group.)
- - - - - - -
If you wish to respond openly to Frank's letter, we'd be happy to publish your reply. We may do a some light editing of any letters, but only where necessary to clarify the meaning.
Chris & Roger
-oOo-
About eComWatch
eComWatch is edited and published by Roger Whitehead and Christopher Ogg. Copyright Roger Whitehead and Christopher Ogg, 2002. eComWatch may be circulated freely in its original format with copyright notice intact. For permission to reproduce any article,